ObjectiveNoncompressible torso hemorrhage is a high-mortality injury. We previously reported improved outcomes with a retrievable rescue stent graft to temporize aortic hemorrhage in a porcine model while maintaining distal perfusion. A limitation was that the original cylindrical stent graft design prohibited simultaneous vascular repair, given the concern for suture ensnarement of the temporary stent. We hypothesized that a modified, dumbbell-shaped design would preserve distal perfusion and also offer a bloodless plane in the midsection, facilitating repair with the stent graft in place and improve the postrepair hemodynamics. MethodsIn an Institutional Animal Care and Use Committee-approved terminal porcine model, a custom retrievable dumbbell-shaped rescue stent graft (dRS) was fashioned from laser-cut nitinol and polytetrafluoroethylene covering and compared with aortic cross-clamping. Under anesthesia, the descending thoracic aorta was injured and then repaired with cross-clamping (n = 6) or dRS (n = 6). Angiography was performed in both groups. Operations were divided into phases: (1) baseline, (2) thoracic injury with either cross-clamp or dRS deployed, and (3) recovery, after which the clamp or dRS were removed. Target blood loss was 22% to simulate class II or III hemorrhagic shock. Shed blood was recovered with a Cell Saver and reinfused for resuscitation. Renal artery flow rates were recorded at baseline and during the repair phase and reported as a percentage of cardiac output. Phenylephrine pressor requirements were recorded. ResultsIn contrast with cross-clamped animals, dRS animals demonstrated both operative hemostasis and preserved flow beyond the dRS angiographically. Recovery phase mean arterial pressure, cardiac output, and right ventricular end-diastolic volume were significantly higher in dRS animals (P = .033, P = .015, and P = .012, respectively). Whereas distal femoral blood pressures were absent during cross-clamping, among the dRS animals, the carotid and femoral MAPs were not significantly different during the injury phase (P = .504). Cross-clamped animals demonstrated nearly absent renal artery flow, in contrast with dRS animals, which exhibited preserved perfusion (P<.0001). Femoral oxygen levels (partial pressure of oxygen) among a subset of animals further confirmed greater distal oxygenation during dRS deployment compared with cross-clamping (P = .006). After aortic repair and clamp or stent removal, cross-clamped animals demonstrated more significant hypotension, as demonstrated by increased pressor requirements over stented animals (P = .035). ConclusionsCompared with aortic cross-clamping, the dRS model demonstrated superior distal perfusion, while also facilitating simultaneous hemorrhage control and aortic repair. This study demonstrates a promising alternative to aortic cross-clamping to decrease distal ischemia and avoid the unfavorable hemodynamics that accompany clamp reperfusion. Future studies will assess differences in ischemic injury and physiological outcomes. Clinical RelevanceNoncompressible aortic hemorrhage remains a high-mortality injury, and current damage control options are limited by ischemic complications. We have previously reported a retrievable stent graft to allow rapid hemorrhage control, preserved distal perfusion, and removal at the primary repair. The prior cylindrical stent graft was limited by the inability to suture the aorta over the stent graft owing to risk of ensnarement. This large animal study explored a dumbbell retrievable stent with a bloodless plane to allow suture placement with the stent in place. This approach improved distal perfusion and hemodynamics over clamp repair and heralds the potential for aortic repair while avoiding complications.